Overclocked FX-6300

[RGone]

Probably Two Pages in bios...

...not having the board and few in here having one for long. We just don't know exactly what pages are needed. But like is decribed of these settings for Channel A and B.

 

[XRogerX]

ok each of these SS is showing how I got in there and what I clicked sry for the flash as it was better with flash so you can see it better

I took a ss of both Slots

 

[RGone]

First thing to do in my mind...

...Wish you had saved image of bios to Usb stick using F12 feature of most new motherboards. Pretty sure they all have it now. Moving on. Channel A pics were hard to read. Image below says what I would do first and noted that both channels must be changed. Not just one apparently in giga.

RGone...

 

[XRogerX]

...Wish you had saved image of bios to Usb stick using F12 feature of most new motherboards. Pretty sure they all have it now. Moving on. Channel A pics were hard to read. Image below says what I would do first and noted that both channels must be changed. Not just one apparently in giga.

RGone...

View attachment 159284

Ok so change it to T2 is what your saying?

 

[Mandrake4565]

Yes

 

[XRogerX]

nope I changed both Slots to T2 AND RAISED the volts to 1.25 and put it to 2200 and still nothing no typing

 

[RGone]

RAISED the volts to 1.25 and put it to 2200 and still nothing no typing

Raise volts on CPU_NB to 1.35 and test. By the way ALL G Skill ram that is tested in AMD systems is tested at 2T and not 1T by G Skill. And is some situations the 2T is even more powerful than 1T contrary to popular opinion. Certainly that is not always the case about 2T vs 1T and is why I always do my own testing instead of going only by popular opinion.

RGone...

 

[XRogerX]

yes I did as you suggested and still nothing , wow I cant believe this m this stinks

 

[Blaylock]

Raise volts on CPU_NB to 1.35 and test. By the way ALL G Skill ram that is tested in AMD systems is tested at 2T and not 1T by G Skill. And is some situations the 2T is even more powerful than 1T contrary to popular opinion. Certainly that is not always the case about 2T vs 1T and is why I always do my own testing instead of going only by popular opinion.

RGone...

I did not know this. Man there's so much to learn, so little time.

 

[ShrimpBrime]

Funny you mention the channel A & B settings for the Giga board. When I started OC'ing my 8350 on it, I couldn't undrstand why it was so unstable, even at 4200MHz. Then it was bam...you have to select each memory channel individually and set your timings in twice. I had always used ASUS before that and with an ASUS board one DRAM timings page would set both channels. I can see this as a common mistake as well. Right after I started figuring out the Giga bios, I had my stability back...until it just eventually got too hot and had to replace it with the Saberkitty.

This is not entirely true. There's many Asus boards that allow timings to be set for one the other or both.

Here's my DDR2 Mainboard model M4A785-M - Not a special board, nor top tier. But does com equipped with some decent OC settings.

ADDITION COMMENT:

when selected Both, I can run different timings on DCT 0 and 1. Usually doesn't end up working out well, but is interesting to play with.

 

[RGone]

I did not know this. Man there's so much to learn, so little time.

Well the part about the G SKill ram is from their Tech Support forum where the 2T thing and AMD has been said more than a few times over a few years by a different tech even.

As for the 1T vs 2T thing it happens sometimes. I think because the ram is highly overclocked and just flat needs the time to recover without sending something to the cpu that the cpu thinks is in error. I only really know that after I saw the situation occurr to me when overclocking some mis-matched ram sets...well I no longer just do but now I test for my ownself. It does not take long with something like MaxxMEM2 to test. I am sure there maybe other useful testers.

RGone...

 

[ShrimpBrime]

when overclocking some mis-matched ram sets...

Where's the OC police phone number? was that ext. 667 or 1333? lol jk

 

[RGone]

yes I did as you suggested and still nothing , wow I cant believe this m this stinks

Look Roger here is where I think we are. Often fooling around with ram timings is one of the fastest ways to get a giant crash and to even windup corrupting the bios of the motherboard. Seen a few, not many, even have to send board off for repair due to memory tweaking/testing. So I generally do not try to teach ram tweaking from the keyboard in a forum. What I might do to my own shett...well that is me and my stuff and I have to absorb the risks. But I generally shy away from tweaking ram in a forum setting.

Up front. I would not have a Gigabyte AM3+ motherboard that I had to buy with my own money as far as I can see into the future. Now that is out of the way. We spend more time trying to figure out some of the crazy stuff that Giga puts in the bios or why their boards do X when we expect Y.

So now we know I have not and likely will not have Giga AM3+ mobo and I do still see some oddity in your bios pictures of the ram settings. One of the settings makes NO sense to me. And I do not have a board to see what the heck is going on. So I hesitate strongly to suggest anymore changes.

CPU_NB being raised is helpful with very fast ram in conjunction with fast cpu speed and with multiple video cards in link. So the real deal is you are not really a heavy overclocker or I don't believe you are heavy into benching. I don't see you with two video cards and 4.5 or 4.6Ghz is not 5.0Ghz and beyond where CPU_NB might be helpful with fast cpu speed.

I suggest you drop the CPU_NB back to 2000 where the board works and then raise you cpu speed back up to whatever was your stable speed you tested with P95 Blend mode for at least 2 hours. I just don't see anything else to do logically with a similar board not in my hands where I would find out what or blow the thing up. Hehehe. Sometimes due the great distances apart in a forum...there just is not a good answer that we can all put our heads around. When I worked for DFI a few years ago, I even have had some of our customers send their crazy acting shett to me or my partner and we would actually test ourselves. About 35% of the time we then sent the board in for RMA but that other 65% of the time it was just some crazy setting a board wanted. Not an everyday thing, but it happens. Move that CPU_NB back down and use the board. Heck it might even be a wonky cpu. No way to know from where we all sit.

Sorry and good luck all at the same time.

RGone...

 

[Alaric]

With my very limited experience trying to OC mis matched RAM , matched PNY RAM , and a matched set of G Skill , on the M4A785 M and M5A99FX Pro , (DDR 2 and DDR 3) , I have never gotten anything but an instant BSOD trying to run RAM at 1T with AMD. BUT , we're talking very limited experience here.

 

[Bishoff]

This is not entirely true. There's many Asus boards that allow timings to be set for one the other or both.

Here's my DDR2 Mainboard model M4A785-M - Not a special board, nor top tier. But does com equipped with some decent OC settings.

ADDITION COMMENT:

when selected Both, I can run different timings on DCT 0 and 1. Usually doesn't end up working out well, but is interesting to play with.

This is good to know mate, thanks!!

With my very limited experience trying to OC mis matched RAM , matched PNY RAM , and a matched set of G Skill , on the M4A785 M and M5A99FX Pro , (DDR 2 and DDR 3) , I have never gotten anything but an instant BSOD trying to run RAM at 1T with AMD. BUT , we're talking very limited experience here.

I ran my amd 965 on a Crosshair 3 @ 1T @ 1600. never had an issue. In fact it is still in service as my HTPC & server and runs 24/7. I still love that little rig.

 

[Alaric]

I ran my amd 965 on a Crosshair 3 @ 1T @ 1600. never had an issue. In fact it is still in service as my HTPC & server and runs 24/7. I still love that little rig.

I'm intrigued! Would you attribute that to the very good mobo , the Phenom II IMC , both ? Something else ? Sorry for the thread jack.

 

[XRogerX]

Well with all that said I do agree with you and your right ty everyone for your time and effort on helping me and just for the heck of it this was mt old system b4 the one I have now lol

http://valid.canardpc.com/2083063
this is the CPU I got to 4.5
this was a PhenomII x955
MSI 790FX-GD70
Tri-fire 4890

but right now im saving up for two new Video cards and a swiftech h240-x

 

[Blaylock]

Well the part about the G SKill ram is from their Tech Support forum where the 2T thing and AMD has been said more than a few times over a few years by a different tech even.

As for the 1T vs 2T thing it happens sometimes. I think because the ram is highly overclocked and just flat needs the time to recover without sending something to the cpu that the cpu thinks is in error. I only really know that after I saw the situation occurr to me when overclocking some mis-matched ram sets...well I no longer just do but now I test for my ownself. It does not take long with something like MaxxMEM2 to test. I am sure there maybe other useful testers.

RGone...

Right. When I'm benching I will generally do the same (1T to 2T) just to see which one gives better results. I have noticed 2T can often give better results and most definitely its more stable. I was never sure if it was the software that liked the 1T or 2T setting or simply the hardware being more stable at that particular setting. Common logic makes me think that there is a universal setting that should make your hardware the "most" stable with any software. This of course is not true.

 

[Mr.Scott]

Being an AMD guy mostly, I run all of my sticks at 1T except for a couple sets of DDR2 that just won't. This spans platforms from Slot A to AM3+. It can be done, it just takes patience.

 

[ShrimpBrime]

Seems AMD likes 1T CR at lower memory frequencies. I'd think it to be difficult to run a 1T at 2400mhz, but nothing is impossible.

1600mhz I always liked 8-8-8-24-32 1T or 2T CR, typically over 933 2T often brings stability when in question.

Some boards clock Ram decent on auto timings. Not entirely sure as to why this happens, it's seldom, but happens. Got 2600mhz from the Formula Z using Auto timing set. Why did that work out? Not sure I even want to question the bios on that particular board, it knows something I don't. Cas 9 and 10 get me around 1100mhz not much more. Cas 11-13-13 I see 2600mhz